Patient-derived xenograft model for uterine leiomyoma by sub-renal capsule grafting

Uterine leiomyoma (UL) or fibroid is a benign smooth muscle tumor of the myometrium with a lifetime incidence of approximately 70%. ULs often require medical intervention due to severe symptoms such as heavy menstrual bleeding and abdominal pain. Although the most common and effective management of ULs is surgical removal, the invasive surgical procedure imposes physical and psychological burdens on the patients. Moreover, the economic burden of UL on health care system is enormous due to the high cost of surgeries. Thus, therapeutic options with long-term efficacy to replace surgical management are urgently needed. For the development of such medical options, reliable preclinical research models are imperative. Ex vivo culture of UL cells has been the primary research model for decades. However, recent studies demonstrated that primary cell culture is not a suitable model for UL research, as primary cultures of ULs mostly consist of non-tumor fibroblasts. Here we describe the protocol for patient-derived xenograft of UL, which faithfully replicates the phenotypes of human UL in situ.


BACKGROUND
Uterine leiomyoma (UL) or fibroid is a benign smooth muscle tumor of the myometrium with estimated lifetime incidences of > 70% [1]. Approximately one-fourth to one-third of women bearing ULs requires intervention due to symptoms, which include heavy menstrual bleeding, pelvic pressure and pain. Currently, the most effective and common treatment for ULs is surgical removal. However, physiological, psychological and economic burdens of UL surgeries on patients and their families are enormous [2]. Moreover, the annual medical cost of ULs in the United States has been estimated 4.1-9.4 billion US dollars [3]. Therefore, it is imperative to elucidate the mechanisms underlying the pathogenesis of UL in order to develop effective non-surgical therapies.
The epidemiology of ULs parallels with ovarian activity: there have been no single case of UL reported in prepubertal girls; symptomatic ULs occur mostly during the ages of 30-40 years; the prevalence increases with age; and the symptoms diminish by the time of menopause with tumor volume reduction. However, the symptoms may continue in menopausal women receiving hormone replacement therapies, indicating that UL is dependent on ovarian steroid hormones. Between two ovarian steroids, 17β-estradiol (E2) and progesterone (P4), E2 was believed to be the primary growth promoter of ULs for decades based on experimental evidence from animal and cell culture models. Meanwhile, studies with human tumors in situ suggested that P4 played a more important role than E2 in the growth of ULs. The direct evidence for the mitogenic effect of P4 on ULs was first demonstrated by a study with this patient-derived xenograft (PDX) model [4], and the findings in the PDX study were consistent with outcomes from the clinical trials of Selective Progesterone Receptor Modulators (SPRMs) for ULs [5].
Recent comprehensive genomic analyses identified recurrent mutations associated with ULs, revealing the presence of at least 4 UL subtypes with unique genetic alterations [6,7]. The most common subtype is MED12 mutant leiomyoma (MED12-LM), which accounts for approximately 70% of all UL cases [6][7][8][9]. MED12-LMs consist of a similar number of MED12-mutant smooth muscle cells (SMCs) and MED12-wild-type tumor-associated fibroblasts (TAFs) [10]. Accordingly, the primary culture of MED12-LMs is not suitable to study ULs, as MED12 mutant SMCs quickly disappear when TAFs overwhelm the culture [11,12]. In cancer research, PDX has become a standard research model to fill the gap between in vitro and clinical studies [13]. Here we describe the technical details of the subrenal capsule PDX optimized for ULs as an ideal alternative to cell cultures. Glass rod with balled tip (< 1 mm 3 ball) made from 9 inch Pasteur pipets. Flame pipets ~20 mm below the tip, pull the tip and further burn the closed end to make a ball [17].  9 Exel International Tuberculin Syringes (FST 14-840-50) 9 Dissecting microscope with LED illumination (e.g., Leica, DMS1000, Buffalo Grove, IL) 9 MultiSample BioPulverizer or single BioPulverizers (BioSpec Products Inc, Bartlesville, OK, 59012MS or 59012N) 9 The Parr 2811 Pellet Press (Parr Instrument Company, Mo-line, IL) 9 Biological Safety Cabinet 9 Fumehood 9 Magnetic stirrer and stirrer bar 9 500 ml beaker 9 125 ml or 250 ml baffled flask PROCEDURE NOTE: Although we expect this protocol to work for all subtypes, this protocol is optimized for MED12-LMs, as this subtype is the most common though the most difficult to make xenografts: MED12-LMs generally contain a smaller number of tumor cells compared to HMGA2-LMs, as the bulk of tumor is made of rigid ECM and often bear a highly necrotic and calcified center. Myometrial control tissues may be processed in parallel.

NOTE:
Sterile/aseptic techniques must be used throughout. Procedure must be performed in a biological safety cabinet.

2.
Using a #22 surgical blade, cut the UL tissues into pieces of < ~9 mm 3 in a 10 mm plastic petri dish, excluding any calcified or necrotic portions as well as myometrium. Keep tissue covered with tissue digestion medium (Fig. 1B).

NOTE:
If necessary due to time constraints, tissue pieces can be cut and stored in culture medium at 4°C overnight. Resuspend cells collected in step15 in collagen gel solution using a pipette with a wide-orifice tip.

18.
With a wide-orifice tip, dispense droplets of desired pellet volume onto the surface of a cell culture plate (6 well-plate is recommended).

19.
Incubate the plate containing the droplets at 37°C in a CO 2 incubator for 15-30 min. After confirming the droplets have solidified, add pre-warmed (37°C) DMEM/F-12 to a depth of 0.5-1.0 cm (~6 ml per well for 6 well plate), and gently detach the cell pellet from the bottom with the tip of a P20 micropipette (Fig. 1E).

NOTE:
The cell pellets in the medium can be maintained at 37°C in a CO 2 incubator over night and grafted on the next day.

20.
Transfer cell pellets to surgery suite for grafting.

Ovariectomy and xenografting
NOTE: While inhalational anesthesia can be used, the mouse position would be constrained to the inhalation tubing position. Therefore, injectable anesthesia, which removes restraints to position, is preferred. As this procedure involves human cells, the grafting surgery must be performed in a biosafety cabinet. We use a Leica DMS1000 digital dissecting microscope, as it allows monitoring the surgery through the cabinet window.

21.
Prepare clean post-surgical cage on a heating pad set to low.

22.
Intraperitoneally inject ketamine/xylazine anesthetic. When the mouse is unresponsive to foot pinching and whiskers unmoving, continue.

23.
Apply analgesics at this time as required by your institution.

24.
Apply eye lubricant, and shave dorsal caudal half of mouse.

25.
Treat the entire dorsal caudal half of the mouse with a disinfectant such as chloroxylenol, followed by 70% alcohol. Repeat thrice.

26.
Transfer mouse to a heating pad set to low.
27. Make a 1.0 cm incision through the skin of the dorsal midline parallel to the spine. Separate the skin and muscle wall laterally by probing with scissors ( Fig. 2A).

28.
Identify the ovarian fat pad through the muscle (Fig. 2B). Make a 2 mm incision perpendicular to the spine just rostral to the ovarian fat, avoiding blood vessels.

29.
Reach through both the skin and muscle incisions with forceps, grasp the ovarian fat, and pull the ovary through the incision to the exterior of the body. Hold the ovary away from the kidney, separating connective tissue if necessary.

30.
Remove the ovary by cauterizing arteries and connective tissues. Or clamp arteries and uterus just below the oviduct with a hemostat (Fig. 2C), then remove the ovary and oviduct with a fine scissors on the distal side of the hemostat. To avoid bleeding, wait for > 1 min before removing hemostat.

31.
Return the uterus into the peritoneal cavity.

32.
With gentle pressure from thumbs and forefingers positioned on the muscle wall, gently push the kidney out through the muscle wall incision (Fig. 2D).

NOTE:
The incision length is in opposition to the longitudinal axis of the kidney, such that the kidney is better maintained outside the muscle wall during the grafting surgery.

33.
Using the soft rounded tips of the Moria iris forceps, gently grip the kidney capsule and lift it away (1 mm) from the kidney parenchyma. Ideally the capsule is gripped at the outer edge of the kidney to optimize space for grafts. Pierce the capsule with one blade of the opened spring scissor (Fig. 2E), and move the blade into the capsule for the length of the blade. Make a single straight incision (~2 mm) along the longitudinal axis of the kidney.

34.
Dip the tip of the glass rod into DMEM/F12 in the dish of cell pellet. Holding one edge of the capsule at the incision with the Moria iris forceps, gently insert the wet tip of the glass rod through the incision between the capsule and kidney parenchyma to form a pocket (Fig. 2F). The pocket with a diameter of the pellet should extend far from the capsule incision but not to the hilus.

35.
While holding open the incision with the Moria iris forceps in one hand, gently pick up a single cell pellet from medium with forceps and insert it into the pocket.

36.
Use the glass rod to gently push the pellet deep into the pocket (Fig. 2G), and then release the capsule. To force the pellet deeper, lightly compress the capsule from the outside with the glass rod.

NOTE:
Additional pockets may be made from the same incision and pellets inserted on the same side of kidney. Multiple incisions in the capsule can be made in different locations, but care must be taken to not disrupt the overall integrity of the capsule.

37.
Holding the muscle incision open with forceps, gently guide the kidney back into the peritoneal cavity.

NOTE:
Do not push kidney with forceps tips, rather use the flat handle end or fingers.

38.
Close the muscle wall incision with absorbable sutures in the simple interrupted pattern (Fig. 2H).

40.
Grasp the skin just rostral to the skin incision site with forceps. Probe apart the skin and muscle wall to form a tunnel from the incision site to the nape of the neck, using another forceps.

41.
With forceps holding the hormone pellet, push the pellet through the tunnel, and deposit the pellet at the nape of the neck (Fig. 2I).
POL Scientific protocol 42. Release the skin and close the incision with wound clips or suture.

43.
Return the mouse to a clean cage on a heating pad. Once the mouse is ambulatory, monitoring can be discontinued and the cage returned to the rack. The mouse should be given analgesic in subsequent days and wound clips removed according to the approved IACUC protocol.

Assessing tumor growth in live mice
NOTE: Measurement of tumor volume by IVIS with transduction of reporter genes, such as RFP, is possible. However, the transduction and selection step significantly reduces the viability of SMCs, resulting in a TAF dominant culture, which in turn affects the growth rate of PDXs. Hence, we assess the growth at 4 weeks after transplantation by live surgery or necropsy, which is explained in the following section. Based on the size of PDXs at 4 weeks, the experimental schedule may be revised as needed.

44.
Follow steps 21-28, looking for the kidney rather than an ovarian fat pad. Visually check the growth of PDXs on the kidney through this incision, or expose the kidney following step 32 (Fig. 3A).

45.
Follow steps 37 and 38 if the kidney was exposed.

Tissue collection
48. Euthanize mouse. To obtain blood, anesthetize mouse with ketamine and xylazine. When full anesthetic depth is reached, perform cardiac puncture; or more preferably, obtain blood through enucleation, as there is less cell lysis.

49.
Spray mouse with 70% ethanol to minimize hair contamination. Cut through the skin and muscle walls, and expose the kidneys.

50.
Grasp the renal vessels at the hilus, and cut the kidneys away from the body.

51.
Transfer the kidney bearing PDXs to a small petri dish with PBS, still grasping by the vessels so as not to damage the kidneys.

52.
Image the grafts on the kidneys with a ruler in the viewer along the x-and y-axes (Fig. 3B). So long as the axes are perpendicular, one axis is set on the greatest diameter available.

53.
For an accurate assessment of height, cut through the center of the graft and kidney parenchyma to obtain a cross section which displays the entire height of the graft (Fig. 3C).

54.
Trim excess kidney tissues using a surgical knife, and process the PDX and surrounding kidney tissues for desirable histological analyses. For the extraction of nucleic acid and protein, remove kidney tissues completely, and freeze PDXs in liquid nitrogen.

NOTE:
Comparable to original human UL, MED12-LM PDXs are well-structured by rigid ECM (Fig. 3D and  3E), making isolation of PDXs free of kidney tissues is relatively easy.

55.
Collect host female reproductive tract, and fix for future reference.
Tumor volume measurement 56. Calculate tumor volume ( Fig. 3B and 3C). Although the PDX grows as an ellipsoid as revealed upon complete extraction, the tumor volume has been assessed as a hemi-ellipsoid, the portion of tumor that extends above the surface of the kidney. Tumor volume = ⅔ πabc: a and b = radius on x and y axes, c = height = (h 1 + h 2 )/4.

NOTE:
Homogenization of MED12-LM PDX is particularly difficult due to rigid texture and small size. Additionally, because ECM accounts for a large portion of tissue mass, the yield of cellular protein and RNA is significantly low compared to other types of tumors. Accordingly, a frozen-crushing technique with the Bio-Spec Pulverizer is the preferred method for extraction of cellular materials. Pulverization of human-derived materials must be performed in a biosafety cabinet. Here we describe the method with an individual pulverizer.

58.
Place pre-frozen tissue in the well of the mortar, and insert pestle.

59.
Pound pestle with hammer multiple times to pulverize the tissue.

60.
Scrape the powder into a microcentrifuge tube with a sterile spatula.

61.
Clean the pulverizer between samples, and repeat as needed.
NOTE: Should the pulverizer thaw, clean thoroughly, and repeat with fresh liquid nitrogen.

62.
Proceed with a standard protein and RNA extraction protocol.

ANTICIPATED RESULTS
PDX prepared with isolated human uterine cells self-organize into tissues with comparable histology to the original tissues [4,18]. If the protocol is appropriately followed, one can expect nearly 100% of grafted UL cells to form ULs of measurable size as shown in Figure 3. If undigested UL tissue transplants are used, the take rate will be significantly lower compared to this cell-pellet protocol, as thick layers of ECM in original UL tissue interfere with angiogenesis [4]. If UL cells are cultured for few days before preparation of cell pellets, the initial concentration of TAFs in MED12-LM PDX may exceed 90%. Nevertheless, E2 + P4 preferentially stimulates SMCs, and thus the tumor should form with POL Scientific protocol an extended growth time.
The hormone condition described above is optimized to achieve fast tumor growth. Thus, the dosage of hormones should be modified according to the purposes of the study. Although PDX shrinks and tumor cells became dormant by hormone withdrawal, the survival of tumor cells in PDXs is hormone independent [4]. Thus, hosts may be left without hormone pellets for days after hormone pellet removal to clear the system. UL PDX response to E2 (and P4) can be detected by expression of progesterone receptor (PGR) (Fig. 4A) [4,10]. The technical details of immunostaining have been described previously [19]. Human and mouse cells can be distinguished by nuclear morphology [18,20]: Mouse nuclei are highlighted by bright granular staining for heterochromatins (Fig. 4B). Furthermore, most cells in PDXs, including vascular cells, are of human origin (Fig. 4C). Human endothelial cells are often detected in the host kidney at the boundary with PDXs (not shown), indicating that LM cells have lower angiogenic activities than host kidney. Hence, a substantial concentration of vascular endothelial cells must be present in the original LM culture for the survival and growth of PDXs, given the contribution of host vascular cells to PDX is none to minimum.  Table 1.